EFFECTS INTRAMOLECULAR AROMATICSUBSTITUTIONS : ISOTOPE
July 5 , 1958
[CONTRIBUTION FROM THE SCHOOL OF CHEMISTRY,
3289
RUTGERS, THESTATE UNIVERSITY]
Deuterium Isotope Effects in Some Intramolecular Aromatic Substitutions BY DONALD B. DENNEY AND PETERP. KLEMCHUK~ RECEIVED DECEMBER 23, 1957 2-(2-Deuteriophenyl)-benzaldehyde (I) has been allowed t o react with di-t-butyl peroxide t o give fluorenone. An isotope effect, k H / k n , of 1.38was observed for this reaction. 2-(2-Deuteriophenyl)-benzoylperoxide (11) has been cyclized by heating to give the lactone of 2-carboxy-2’-hydroxybiphenyl(IIIa,b). An isotope effect of 1.32 for k B / k n was found for this rraction. Irradiation of 2-(2-deutcriophenyl)-phenyldiazomethane( V I ) gave fluorene. The isotope effect for this cyclization was found to be 1.12.
In another paper? we have reported some deu- mixture indicated an isotope effect of 1.32 for the terium isotope effects obtained during the acid- lactone formation. catalyzed cyclization of 2-deuterio-2’-carboxybi- Recently Walling5 has discussed free radical subphenyl. I t seemed of interest to extend our studies stitutions in aromatic systems. He has considto some other types of reactions. In this paper we ered three mechanisms A, B and C. On the basis wish to report results obtained from two freeR. ArH -+ RAr H. (A ) radical substitutions3and a carbene substitution. R . When 2-(2-deuteriophenyl)-benzaldehyde(I) was R . + ArH +RH + Ar. +RH (B) heated with di-t-butyl peroxide for thirty hours it R . cyclized to a mixture of fluorenone and deuteroR. ArH --+ R A b H+RAr RH (C) fluorenone. of the energies involved for these paths and the prodD ucts obtained, he has considered mechanisms A and B as being very unlikely and has suggested that path C is the most reasonable. Our results are in \ agreement with his interpretation since both A HC=O and B would most probably lead to rather large I isotope effects in the cases reported here. It is not D necessarily true that A and B would always show I large isotope effects since this will depend on the amount of C-H bond breaking in the transition states. With relatively unreactive radicals such as the carbonyl and carboxyl radicals one predicts that the C-H bonds will be significantly weakened in the transition states and therefore a large isotope effect should be found. With very reactive radiA cals this might not be the case. + Mechanism C can be used to explain the rather small isotope effects found in this work. If in the cyclization of I and I1 an intermediate I V is
+
+
+
D
/ ~
0-c
/
I/ IIIa O
\
x
X
-
\
+
~
\
/ 0-c=o IIIb
Deuterium analysis of this mixture indicated an isotope effect, k H l k D , of 1.38for this reaction. When 2-(2-deuterophenyl)-benzoylperoxide (11) was heated in carbon tetrachloride, according to the procedure of Kenner, et al.,4i t gave a mixture of the lactones, IIIa and IIIb. Analysis of this (1) Alfred P. Sloan Fellow in Chemistry, 1956-1957. ( 2 ) D . B. Denney a n d P. P. Klemchuk, THIS JOURNAL,80, 3285 (1958). (3) Recently i t h a s been reported, C. C. Price a n d R. J. Convery, ibid., 79, 2941 (19571, t h a t no isotope effect was observed during t h e phenylation of a mixture of 2,4-dinitrobenzene a n d 2,4-dinitrotrithiobenzene. A later communication, C . C. Price, i b i d . , 79, 6579 f1957), refutes their earlier findings a n d they now report t h a t , “Because o f t h e low conversion, no conclusions concerning a n isotope effect can be made.” (4) G W. Kenner, hf. A. M u r r a y and C. M. B. Tylor, Tetrahcdvon, 1. 259 f 1 9 3 i ) .
Droduct
V IV, x = co, CO? formed and this can return to V or a a-complex derived from V a t a rate which can compete with the conversion of IV to products, then one will observe an isotope effect, providing there is an isotope effect in the conversion of IV to the products.6 The mechanism by which IV is converted to products is not completely understood; however, one would expect an isotope effect for this process whatever it may be. There seems to be no compelling reason for not accepting an equilibrium between IV and V or the a-complex in these cases since both the carbonyl and carboxyl radicals are relatively stable species. A less stable radical might not return and therefore one would not observe an isotope effect even though the mechanism remained the same, +
~
-
( 5 ) C. Walling, “Free Radicals i n Solution,” John Wiley a n d Sons, Inc., New York, N. Y.,1957. p. 482. (6) H. Zollinger, Helv. Chim. A r l a , 38, 1597 ( 1 9 5 5 ) , h a s derived t h e kinetic expressions for t h e entirely analogous case of electrophilic substitution.
~
3290
DONALD
R . DENNEY AND PETER P.
There is another way in which these isotope effects could arise; in the formation of IV if there is any radical hyperconjugation of the carbon-hydrogen bond a t the carbon being attacked then k H / k D will be greater than one. In our first paper we discussed this factor in detail. I n these cases it seems somewhat unreasonable to attribute the isotope effects observed to radical hyperconjugation ; however, i t cannot be eliminated as a possibility. When 2-(2-deuterophenyl)-phenyldiazomethane (VI) was irradiated it gave a RlSI, yield of fluorene (VII). D
( D or H)
o-bAa-'a-+ \ + H-CES=~Y \/ A
( D or H)
VI1 13
VI
KM1104
\/ I! 0
In order t o determine the isotope effect for this reaction i t was necessary to oxidize the fluorene to fluorenone, which was then analyzed for deuterium ; using this procedure k H / k D was found to be 1.12. A reasonable mechanism for this reaction involves formation of the carbene VIII, which then reacts to
--(_I=. /=\ / \ --+ \C
H /
VI11
KLEMCHUK
VOl. 80
made basic with 1Oyo sodium carbonate solution. The aqueous solution was concentrated t o remove acetone treated with Xorite A and acidified t o yield 0.274 g. (81.4 of acid, m.p. 110.0-113.5°. The acid was purified by cr tallization from acetone-hexane (1 :9) to give 0.252 m.p. 1 1 2 . f ~ 1 1 4 . 0 (lit.* ~ 113.5-114.5°). The acid was analyzed for deuterium; found: 9.88 =t 0.05, 9.73 f: 0.05 atom c/o D. Cyclization of I.-Compound I(1.83 g., 0.010 mole) mas heated under nitrogen with 1.00 ml. of di-t-butyl peroxide in an oil-bath at 143 f:3'. After 17 hr. an additional 1.00 ml. of peroxide was added and the reaction was continued for 13 hr. The solution was eoncentrated in z'acuo and then evaporatively distilled to yield 1.33 g. of a clear yellow liquid. The distillate was chromatographed to give 0.89 g. of recovered aldehyde and 0.37 g. of fluorenone. The fluorenone was purified by crystallization from hexane, n1.p. 83.&85.0". Anal. Found: D, 7.08 i 0.04, 7.21 f 0.04, 7.13 k 0.04. A portion of the recovered aldehyde was oxidized to the corresponding acid, m.p. 112.0-114.0°. This material was analyzed for deuterium. This result shows quite clearly that no exchange took place during this reaction. Anal. Found: D, 9.78 f 0.06, 9.86 =k O.0F. Cyclization of 2-(2-Deuteropheny1)-benzoyl Peroxide.The procedure of Kenner, et ~ l . was , ~ followed. From 1.00 g. of 2-(2'-deuterophenyl)-benzoie acid there was obtained 0.788 g. of peroxide, 1n.p. 107-109" (lit.3 107-108'). This material (0.548 9.) was Converted to 0.067 g. of lactone, m.p. 58-91", by Kenner's method. After further purification by crystallization from hexane there was obtained 0.025 g. of lactone, 1n.p. 93.0-94.0' (lit.33986-88', 92.5"). This material was anall-zed for cleuterium; found: 7.(Jl f 0.04 atom yo D. Preparation and Cyclization of 2-(2-Deuterophen phenyldiazomethane .--2 - (2-Deuterophenyl) - benzaldeh (0.910 g., 0.00496 mole) was refluxed for one liour \ 0.590 g. (0.010 mole) of 85';1, aqueous hydrazine hydrate in 10 nil. o f absolute ethanol. T h e solution was evaporated in wczm to yield a n oil which was dissolved in 2
